Assisting Community-Based Oncologists and Surgeons in Making Treatment Decisions for Patients with ER-Positive, HER2-Negative Early Breast Cancer
Assisting Community-Based Oncologists and Surgeons in Making Treatment Decisions for Patients with ER-Positive, HER2-Negative Early Breast Cancer
Video proceedings from a roundtable discussion with Drs Harold J Burstein and Tari King featuring discussion of a comprehensive case-based survey focused on decision-making for patients with ER-positive, HER2-negative early breast cancer, actual patient cases, relevant data sets and ongoing clinical trials.
2.25 AMA PRA Category 1 Credits™
To complete the Post-test and Educational Assessment and Credit Form, click here. A statement of CME credit will be issued for scores of 80% or better. Your statement of credit will be mailed to you within 3 weeks or may be printed online.
Overview of genomic assay use to assist in clinical decision-making
DR BURSTEIN: The first thing is that your data show that in 2017, breast cancer is living in the genomic era. Everybody is getting these tests. Working oncologists, academic specialists, we’re using these tests every day. And it is a vocabulary that people have gotten familiar with and warm to. And there’s still this kind of straw man thing. Speakers put as their slide, like, “Remember when we used to give chemo to everybody?” But, I mean, that’s 20 years ago. And for the past 10 to 15 years, ever since the Oncotype data came out in 2003/2004, we’ve really been shifting towards this. And what you’re seeing here is totally predictable, that there’s widespread use of these assays. And I think the market data that you reflect here probably fits as well. The most common test is the Recurrence Score®. There are a variety of other products that are coming along, which mostly echo the same themes that have been articulated very nicely in the Recurrence Score literature. DR LOVE: Another thing, Tari, is the issue of the surgeons being involved in the decision and actually kind of knowing how it’s going to play out with their oncologists in terms of what assay to order in what situation. Can you talk a little bit about how you’ve interfaced with medical oncologists and the interdisciplinary team both when you were at Memorial and now at Dana-Farber? DR KING: Sure. Yes. So breast cancer, obviously an incredibly collaborative, multidisciplinary specialty. And for many of our patients, particularly the ones that we’re talking about here, the ER-positive, HER2-negatives, as Hal suggested, the decision about using therapy has shifted so much towards these assays. And so it’s really important for us to work together as a team and know when we want to use one and when we don’t want to use one. Also, timeliness of care is a very important issue. And so we know that it takes some time to get our patients’ material sent off for these assays. It takes some time to get it back. And so what we’ve done together, both at Memorial and at Dana-Farber, is to develop some criteria where the surgeon can recognize when the pathology report’s available. We look at the report, and we know which patients our oncologists are going to want the information from the test and which ones they’re not. And so we will go ahead and initiate that testing, again, really improves the turnaround time to get the result back and keeps us all on the same page as far as how we’re making decisions about patients. DR BURSTEIN: Tari’s being modest here. But I want to point out, this is a way that surgeons can really help out an oncologist, because we didn’t have a pathway for this. Memorial had implemented a pathway, led by the surgical team. If it’s this size and this appearance, we’re going to order this test because we always get it. And under her leadership, we’ve rolled that out at Dana-Farber as well. And it has dramatically shortened the turnaround time for getting the genomic information back. And it’s a pretty fixed algorithm. Occasionally we forget it’s there. And you’re looking for it, and it doesn’t show up. And everything falls through the cracks once in a while, but it’s been a real work-flow improvement for our team. DR LOVE: When you went through this process, Tari, what did you end up with in the algorithm in terms of which assay in what situation? DR KING: As your data reflect, we do primarily rely upon the Recurrence Score. It’s been available the longest. Physicians are most comfortable with it. It’s in the NCCN Guidelines. And it does give us not only the prognostic information but the predictive information about response to therapy. And so that is really what we rely on for our routine ordering. And when we set our criteria for when the surgeon would initiate the ordering, it was a combination of things that we discussed with our medical oncology colleagues. But it comes down to the traditional things like tumor size and whether lymph nodes are involved or not. At Dana-Farber, our oncologists, they do pay a lot of attention to the grade of the tumor, so grade was a factor. We do not order on anybody with Grade 1 disease. And then there’s some other considerations as far as patient age and what not, but generally very simple criteria where the surgeon will know when the test is going to be useful for the medical oncologists. And it also, again, helps the surgeon in talking to the patients, because the patients are asking us these questions when we’re giving them their pathology report, too. “Am I going to need chemotherapy? What do you think?” And this provides a framework for us to have that discussion and to not contradict something that our oncology colleagues might say.
Role of clinical characteristics in adjuvant therapy decision-making
DR BURSTEIN: Overall, there’s a very broad concordance between things like grade and proliferation, or Ki-67, and genomic assays. So lower-grade tumors tend to have lower rates of proliferation. Strong ER and PR expression, they tend to have lower genomic indices, whether it’s luminal A characterization or a low-end Recurrence Score. And at the other end of the spectrum, it all goes the other way. So in the broad measure, these all make sense. At any given site, I think people value different things. We feel very lucky to have superb pathology. And we do put a fair amount of stock in grade assessments. But even our pathologists would acknowledge that they don’t have a lot of data that their grade assessment really predicts what happens in the long term for any given patient. Ki-67 is a very controversial marker. It’s probably a pretty good test. There have been dozens and dozens of evaluations of Ki-67. And the thing about Ki-67 is, every study that’s ever been done shows that higher proliferative scores do less well than lower proliferative scores. It’s an unbelievably consistent finding. The issue is that when you compare center to center or pathologist to pathologist, it’s very hard to know where to set the benchmarks. So the example that I use is, if you have a 4.0 at MIT, you’re a smarter kid than if you have a 2.8 at MIT. But you don’t know how a 2.8 at MIT compares to a 4.0 at Mass Bay Community College. And that’s the dilemma in terms of figuring out where to set the benchmark for Ki-67. It’s very hard to know whether their assessment is the same as their assessment. And it’s been hard to tease that out in the literature. So if the story all fits, these are all useful. But you do begin to see, when you look at individual cases, where the granularity breaks down the relationships. DR LOVE: Tari, what about the histopathology? Obviously, most of the data we have is on adenocarcinomas, infiltrating ductal adenocarcinomas. But, for example, lobular cancer, there are other subtypes. How does that affect the way you look at these patients? DR KING: Mm-hmm. It’s a very important question, because we do know that these scores were generated largely from patients with invasive ductal cancer, as you suggested. I also wanted to point out, it’s interesting if you compare what your surgical oncologists rely on here in your slides compared to your medical oncologists in the slide before. You can see the surgeons are a little less confident in the grade and the Ki-67 than perhaps the medical oncologists. DR BURSTEIN: Skeptical. DR KING: I don't know what that says about us. Yes. But yes, we do — lobulars are a very unique subtype. We’re learning more and more about them. I tend to really go back to what we largely know, is that they are very estrogen receptor-dependent cancers. And so instead of focusing so much on lobular versus ductal, I really think that in general, these are strongly driven by estrogen. And I do rely on the genomic assays when discussing the lobular cancers. One of the unique things, though, about the lobular cancers is that we can see these patients with multiple positive nodes. And so it makes us a little anxious, because we have the woman with otherwise what we would consider a good breast cancer, but yet she has multiple positive nodes and we’re not sure. Our gut says, “Gosh, do we need to be more aggressive?” But the genomic assay says it’s driven by estrogen. And so I think those are challenging, but the more experience that I have with these assays and with patients and seeing these assays also used in different clinical settings — because I have done a research project where I’ve actually performed some of these assays in women with metastatic disease, doing the assay on the primary tumor but the women had metastatic disease at the time. And we see that these assays remain very, very prognostic even in that setting. And so I think it’s not so much about anatomy but really is about biology. And so regardless of the histologic subtype, regardless of the stage of disease, I think these assays have — at least the Recurrence Score, the one I’ve examined, they really do seem to hold up. DR BURSTEIN: That’s a really good point. Now, across the spectrum of preoperative therapy, adjuvant therapy and even metastatic treatment, you’re exactly right that the genomic prognostic model really is quite robust. The numbers vary in the different settings, but there’s a clear lesson there. DR LOVE: Although, I mean, I don’t know that people look at genomic assays that much for practical purposes in the metastatic setting, but kind of you wonder whether they should. Do you ever use genomic assay information in the metastatic setting? DR BURSTEIN: To be honest, most of the time we actually have that. In the real world it’s because the test was ordered, and then it turns out they had metastatic disease. Tari has published work on looking at the Recurrence Score as a predictor of outcome in metastatic disease. And perhaps predictably, but the lower Recurrence Score cases tend to have a longer progression-free survival, tended to do better with endocrine therapy — I forget exactly the details — but needed chemotherapy less soon than did others. And so again, I think it’s a story that holds together across the stages of breast cancer. DR LOVE: Any other comments about your work? DR KING: Yes, just to be clear. I mean, it was an exploratory study within the concept of a larger study where we were looking at the role of surgery in patients presenting with de novo metastatic breast cancer. But the findings were very intriguing in that it did suggest the hypothesis that if you presented with Stage IV disease and you had a high Recurrence Score that perhaps you should receive chemotherapy first as opposed to the traditional hormone therapy in metastatic setting. And there is some discussion going on within the Alliance, the clinical trials program now as to whether we could design a Phase II study to actually test that hypothesis. And so we may be seeing that coming along.
Effect of tumor size and nodal status on the decision to administer adjuvant chemotherapy to a younger patient with ER-positive, HER2-negative BC
DR LOVE: We presented a scenario of a 40-year-old premenopausal patient, 65-year-old and 75-year-old patient, all with ER-positive, HER2-negative disease. And we summarized what we saw in the younger patient here, which is on the left, you see people with positive nodes, from 1 going up to 4. On the right, you see node-negative cases. And these are the medical oncology investigators. So in a 40-year-old patient with node-positive disease, actually the majority of investigators would use a genomic assay. And you can see down below that it’s almost all the 21-gene Recurrence Score at this point. That’s with microscopic involvement. You still see three quarters of them using it with 1 positive node, starts to drop down, 50. And it goes kind of what you would think. You look at node-negative, almost universal use of, again, Recurrence Score. And pretty much, I think, very similar answers coming from the surgeons, although not identical. But Hal, I guess the first thing that struck me — and if I had to pick 1 reason that we did this survey, I was just very surprised that the ASCO Guidelines statement about not using a genomic assay in node-positive disease, when we know — we were asking the noted investigators, and they were using it, particularly for 1 node. And we wanted to just be sure we were right. Looks like we are. Now, I don't know politically kind of what it means, but at least it’s a different set of information for an oncologist or physician or patient to think about. Any comments? DR BURSTEIN: Across a whole spectrum of stage, the general principles really seem to hold. Lower genomic assay results tend to have a better prognosis with endocrine therapy, higher a less good prognosis, probably warranting chemotherapy. Everything that we’ve learned suggests that that would be true for node-negative and for node-positive. And historically, we started using these tests in the node-negative patients because we had very clean data from the analysis of NSABP-B-20, and also because the risks overall were smaller. Patients who have node-negative disease tend to do pretty well. And so you scratch your head and you say, “What if I’m really wrong? What if this assay leads me down a path that’s a big mistake?” In general, you’re not making a catastrophic error, because the benefits of chemo are still going to be pretty small in women who have smaller node-negative tumors. The quality of the data for use in node-positive is exactly the same as it is in node-negative. There was a very well-done study presented by Kathy Albain, SWOG-8814. She looked at the Oncotype DX 21-gene Recurrence Score as a predictor of benefit from chemo. And qualitatively, the results were the same. That is to say, low-risk scores didn’t seem to benefit from anthracycline-based chemo. Higher-risk scores did. We didn’t incorporate that into the first wave here, because people were worried. “What if we’re just overselling this? You don’t want to miss somebody who’s got a higher-risk node-positive. As the data have now come forward over the past 10 to 15 years, I think the comfort level with using these tests in the node-positive cases has gotten more and more, particularly in the 1 or 2 positive nodes, which, of course is all we see, because our surgeons are only taking 1 or 2 nodes these days. And I think, as your survey suggests, the vast majority of oncologists and surgical oncologists are very comfortable looking at these tests for lower-risk node-positive cases as well. The ASCO Guideline did not recommend that. There are other guidelines, including NCCN, that have been more forgiving about that. And so it usually is covered by third-party payers and insurance. And we certainly do it all the time on a day-to-day basis in our patients. DR LOVE: Any comment on the issue of the fact that most of these patients don’t have an axillary node dissection? So even though you see an algorithm that’s driven by number of nodes, there’s a certain level of uncertainty in terms of what the true number might be. DR KING: That’s a very interesting point. I mean, as the surgical community has pulled back in our extent of surgery because we’ve demonstrated that more surgery is not better for most patients, it does call into play these decision-making tools where the number of positive nodes are critical. But again, I think if we go back to that principle that we’ve been now stressing, it shouldn’t really make a difference, the number of positive nodes that you have, if you’re really talking about the biology being driven by the features of the primary tumor. DR LOVE: Any thoughts about the 70-gene assay? And we have the MINDACT data out there. Any potential advantages? One thing we have heard people talk about is the issue of patients with node-positive disease. Do you think it presents any potential advantage? DR BURSTEIN: So as the literature has evolved, we’re now getting experiences with node-positive patients, including some of these that were in the MINDACT study. And we also have some data from the West German PlanB study and some other experiences looking at the Recurrence Score. So I think that the data are now catching up to what the doctors are doing, which is always a good thing. The issue with MINDACT is, yes, it included some node-positive patients. So I think it extends the ability to use these genomic assays into those patients and the same overarching principles still apply. Having said that, it’s still in the lower end of the nodes. And just to follow up on Tari’s point, I think we all are getting better at trying to think about what sentinel node information means. So, I mean, if you have a big chunky single node and there’s only 1 taken out, that’s probably different than someone who had 3 sentinel nodes taken out and there was a little tiny, less submillimeter focus. And the Memorial and other hospital algorithms have helped us think through, “How likely is it that there really is a lot of other disease that we’re not aware of?” Every once in a while you worry about that, because if you say, “There were 2 out of 2 nodes. But if there were 14 positive nodes, yes, I’d give them chemotherapy.” And we just don’t really know. But, fortunately, that doesn’t happen too often.
Importance of patient age in adjuvant therapy decisions, including the use of genomic assays
DR LOVE: The next thing I would like your input on is the same exact case scenarios but now increasing the age, both age 65 and 75. And if you think about what you might expect, I think that’s kind of what we see. So here’s the 40-year-old. We move to a 65-year-old, and you see a lot more yellow. Particularly in the node-positive patients, a later trigger, so to speak, more nodes, people thinking about, again, mainly the 21-gene Recurrence Score. And on the lower end for the older patient, now in patients with tumors that are smaller, particularly under a centimeter, just saying, “Hey, I’m not going to be using chemotherapy, so we don’t need to do anything.” And again, surgeons, pretty much the same thing at age 65. Hal, any thoughts about this? DR BURSTEIN: It both makes absolute sense and it probably misses an important point. So the absolute sense part is, historically we’ve known that chemo is a lot less important for older patients, especially a 65-year-old and over. But that’s probably because they get better-biology cancers, to a large degree, and they’re much more likely to be ER-positive, PR-positive and HER2-negative. And so the interesting irony is that as you normalize for biology, it may not be so clear that 65-year-olds don’t get the benefit. What does happen as you get to 65 — and every clinical team knows this — is that there’s much more heterogeneity, on average, in the other health problems. There are some 65-year-olds who’ve got 25 more years in them. They can tolerate chemotherapy. And there’s a growing number who are not so healthy, will struggle with chemotherapy, where it’ll have much more of a negative impact on their quality of life or where other health conditions are going to loom larger in the next 5 to 10 years for that patient. So that clinical heterogeneity really makes everyone more ambivalent about recommending chemotherapy across the board. DR KING: And patient preference comes into it a lot with the older patients also. The younger woman may accept a therapy for a very small benefit. And the older woman may feel a little bit differently about that. DR LOVE: And I think this trend, actually, you see it even more so as you increase the age. And you start to get into the issue of, Hal, of how old is too old to get adjuvant chemotherapy? I mean, if I were to present a 95-year-old, I think we would all agree. But in a completely otherwise healthy patient, Hal, would you consider adjuvant chemo at age 80, at age 85? In any event, we chose to ask about 75. And you can see, again, more red in the node-negative, so less of a trigger to consider chemotherapy in the lower-risk patient and more genomic assays. Hal, any thoughts about this? And again, we see, I think, pretty similar things with the surgeons about the older patient. And again, how old is too old for adjuvant chemo? DR BURSTEIN: Seventy-five is getting close to being too old for ER-positive breast cancer — HER2-negative breast cancers. You’d have to have a really healthy 75-year-old and a really high-risk cancer to make you want to get chemotherapy in those circumstances. One of the frustrations has been that our chemo regimens haven’t really changed much in the past 20 years. Ever since Larry Norton gave us dose-dense AC/paclitaxel and ever since the US Oncology group came up with the TC regimen, that’s been about it. And those are tough regimens for 75-year-olds. As I tell my septuagenarians, chemo has a way of making a very vigorous 70-year-old feel like they’re not so vigorous and they’re about 80. So I think if we had better-tolerated chemo, it actually might shift this curve. The problem has been it’s hard to find that. And Hy Muss did a very nice study in the CALGB. And the trouble was that the easier chemo just didn’t work as well. DR LOVE: That was capecitabine. DR BURSTEIN: And that was also true — that was capecitabine. And then Larry Shulman did a study in the CALGB for single-agent paclitaxel. And that didn’t work quite as well either. So the trouble has been that to get the benefit of chemo, you so far have to really get a more intensive chemo regimen in. For many 75-year-olds, that just doesn’t add up. DR LOVE: I guess, Tari, when you really get down to it, the thing I kind of think about is the patient who has a genomic assay that’s high-risk. DR KING: Mm-hmm. DR LOVE: And in an otherwise healthy patient, seeing the potential benefit even in people older than we might think, if they were presented with it, maybe a fraction would want it. Any thoughts? DR KING: Yes, I agree. I think that we live in a time with a lot of shared decision-making, patient-centered care. And for women who do have not our typical 75-year-old woman cancer — so again, going back to most 75-year-olds get cancers that we feel very comfortable treating with endocrine therapy. But when you get an outlier, I do think that the more information that you can have to discuss with that patient, what the potential benefits/what the potential risks are, it’s very helpful. Because women do want to be involved in the decision-making and, even if a woman decides not to take the more aggressive therapy, it’s important that she have a realistic expectation about what the outcome may be. DR BURSTEIN: And there, too, that’s where these genomic assays have been transformative. Neil, you’ll remember this study that the Mayo group did, like, 25 years ago. And they had the surgical and medical teams estimate the risk of recurrence of the breast cancer and then the likely benefit of chemotherapy. And the numbers were all over the place. DR KING: Mm-hmm. DR BURSTEIN: How could a patient make a good decision when there was so much absolute guesswork? Now, with these assays, if you integrate stage and genomic information, you can give a pretty good ballpark about, “Here’s what the risk is. And here’s how far the needle moves when you dial in the chemotherapy.” And most patients can look at this kind of information and say, “Yes, that makes sense,” or, “What? That’s all you’re talking about? That’s not worth it to me.” And particularly the older patients, I think, are very capable of doing that.
TAILORx: Results from the low-risk registry of a prospective trial of adjuvant systemic therapy based on the 21-gene Recurrence Score® (RS) for patients with ER-positive, HER2-negative BC
DR BURSTEIN: So TAILORx is an NCI-supported study being led by Joe Sparano at Albert Einstein. And it was a study for ER-positive, node-negative breast cancer patients where, as part of their initial evaluation, they had the tumor tested for Oncotype DX Recurrence Score. And then the treatment was stratified based on that result. If the score was low, defined as less than 11, they got endocrine therapy alone. And if the score was high, greater than 25, they got chemotherapy and endocrine therapy. And in between, there was a randomization. So the goal here was to be a prospective validation of the idea that the Oncotype DX could stratify patients into treatment groups. And the only difference between this and the conical studies for Recurrence Scores is that they shifted the goalposts so that the Recurrence Score cut points were 11 and 25 instead of 18 and 30. And they did that because they were really trying to define where in that intermediate zone you would find the inflection or the likely benefit for chemotherapy. So what Joe published a year or 2 now ago in The New England Journal of Medicine were the first outcomes from this trial, which was a prospective look at the very low group, so scores of 10 or less. And it made the point that these patients have an extraordinarily good prognosis, so a major paper in The New England Journal of Medicine. Most patients had Stage I, a few had Stage II breast cancer. There was a distribution by grade. And if you look at the inset boxes — it’s always a good thing when The New England Journal has to publish a little inset box that focuses on the range between 90% and 100% cancer free. So that’s a good problem. But if you look really closely, you’ll see that you’re talking about a 5-year risk of 5% or less with endocrine therapy alone. So vide infra, there is no possibility that chemo could have helped these people do better. And it validates the habit we’d already gotten into of using endocrine therapy alone for these low genomic assays. DR LOVE: Tari, any comments about this really classic paper, a prospective study looking at the same thing that we’d already decided a while back based on retrospective prospective kind of analyses? DR KING: Mm-hmm. Yes. No. I think it was very nice to see the low event rates in this group, reaffirming what we’d been doing. One of the challenges, I think, with this study was that they did change the cutoff points for the low risk and the high risk. And that left some oncologists a little bit unsure about how to use the test in everyday practice. If they got a score of 15, is that low, or is that intermediate? And we saw that some in our practice. But this is very reassuring, and we look forward to the rest of the results. DR LOVE: And when do you think we are going to see them? Of course what people want to see is the chemo versus no chemo, but also the companion study, the RxPONDER in node-positive disease. What’s that looking at, and when are we going to start to see these data come out? DR BURSTEIN: You’ll probably have retired before we see these data coming out. The event rate in these studies is unbelievably low. It speaks to this shift that we alluded to earlier about how very well lower-risk patients in contemporary practice are doing. And if you look at distant metastatic recurrence in the TAILORx study, you’re talking about less than a 3% risk at 5 years. And other studies have shown similarly good trends. So in the middle group, this cohort where there’s the randomization, it’s driven by how many events there are. And there are hardly any events. So I don't know exactly when it’s going to be. Increasingly, the guess is that it’ll be negative, that is to say that all of that cohort between 11 and 25 is going to do very well regardless of chemo, because it may be that the patient selection factors that doctors implement are so good that when you take smaller tumors with better grade and lower Oncotype, you just can’t see much of a benefit for chemo even up to a score in the midtwenties. So obviously it’s a really important question, though the more I’ve thought about these data, I think the question has actually shifted. This was to make sure chemo didn’t help low scores. Now what we want to do is go to the kind of cases we were talking about earlier today, taking the Stage III cancers which have low scores and seeing if they need chemotherapy. That might be more productive, because at least there are going to be more events to answer the questions.
RxPONDER: A Phase III trial of adjuvant endocrine therapy with or without chemotherapy for patients with node-positive BC and a RS of 25 or lower
DR KING: The adoption of these genomic tests in node-negative patients was certainly earlier on and, I think, a little bit easier for people to digest. And the adoption in node-positive came much later. It’s just unfortunate that we couldn’t just do 1 trial with everybody in it and that we have to wait for the 2 trials. But the RxPONDER trial, I think, certainly accrued very rapidly, because physicians were interested in this question. But based on what we’ve learned since the time that that trial started, my gestalt would be that the results are probably going to be very similar. I mean, I don't know if you feel differently. But again going back to this, that it’s really the biology not the anatomy that’s driving it, I mean, the event rates will probably be higher in the node-positives. But I don't know that we’ll see — I still would suspect that we’ll see that the low-risk group does perfectly fine without chemotherapy. And where it falls out in the intermediate, I think, will be the most interesting part. DR BURSTEIN: The other challenge is that this is a disease where recurrences readily happen after 5 years. In fact, more recurrences happen in years 5 through 10 — 5 through 15 than happen in years 0 to 5. And 5 years is probably less than half of the events for most cohorts. So when you’re looking at data, even powerful data like these where you’re only looking at 60-month endpoints and the median follow-up is less than that, you haven’t seen most of what you’re going to see. The downside is, it takes a long time to accumulate 10 years’ of median follow-up. And it’s just going to have to play itself out. DR LOVE: And the RxPONDER trial allowed up to 3 nodes. DR BURSTEIN: That’s my recollection. DR KING: Yes. DR LOVE: And what was the thinking there? We would think biology is biology. DR BURSTEIN: Biology is biology. I think it’s easy to look back at this and other experiences and say, “Oh, they set the gates too low. It should have been a higher score.” You don’t want to get something big wrong. And I think that the transformative nature of genomic testing has been hugely influential. We’re now debating the extreme cases of it. That’s not the meat and potatoes of most early-stage breast cancer in the United States. And it’s important to get those data. There’s been a small-c conservatism that people didn’t want to quite move too fast. Maybe we should have been a little bolder. But I think you have to also respect the sense that if we’re withholding a treatment like chemo, which had been given for 20 to 30 years and had a proven survival advantage in node-positive breast cancer, before we just chucked that, you wanted to be really sure.
Results of the Phase III West German Study Group PlanB trial: Significance of the 21-gene RS and concordance of prognostic markers in high-risk ER-positive, HER2-negative BC
DR BURSTEIN: So this was a prospective study, very similar to the TAILORx study. And the report here is on the outcome of women who, again, had very low Recurrence Scores, in this case less than 12, who only were assigned to receive endocrine therapy. And what this group showed with even less follow-up now at 3 years, but was again a superb prognosis for node-negative and node-positive patients, a 98% 3-year disease-free survival, and so again making the argument that there’s no way chemotherapy would have helped these patients. The novel wrinkle here is 2 things. One is that they included some node-positive patients, so it’s 1 to 3 positive nodes, and again, an excellent short-term prognosis. And the other is that they actually began to look at some of the slightly higher Recurrence Scores. And again, those patients are also doing really well, though perhaps not quite as well as the extreme low scores. So mostly a prospective validation, very similar to TAILORx, but including a few node-positive patients.
MINDACT trial: Utility of the 70-gene signature in selecting patients with BC and 0 to 3 positive nodes for adjuvant chemotherapy
DR BURSTEIN: So MINDACT was a fantastic collaboration amongst a huge number of European-based investigators. And the fundamental question here was, what is a better decision-making aid? Is it clinical stage, or is it a genomic assay? Or is it some combination of the two? And fundamentally, the answer is that a combination of the two is the best approach to thinking about treatment decisions for the kind of patients we’re talking about today, ER-positive, HER2-negative breast cancer. The operation of that study was really complicated. So what happened was, a woman had surgery. The tumor was then staged by traditional pathology up to 3 positive nodes. The study did include ER-negative and even some HER2-positive breast cancers, though the numbers ended up being so small I don’t really think it informs those treatments. The tumors were then also sent for the MammaPrint assay, which categorizes tumors as low or high risk. And then they were put onto the study with a treatment algorithm that looked like this. If the clinical risk was low and the genomic risk was low, they just got endocrine therapy. If both were high, clinical and genomic, then they all got chemo. And in between, if there was a discordance between the clinical and the genomic risk, there was a randomization to endocrine therapy with or without chemotherapy. Now, that’s when the wrinkles begin. So how did they define clinical risk? They actually used the Adjuvant! Online program that Peter Ravdin had developed. They plugged in the variables about grade, tumor stage and ER and HER2 status. Unfortunately, that is now offline. You can’t use that anymore. So they have printed in their appendix to their New England Journal of Medicine publication a grid that allows you to figure out who’s at clinical low risk and high risk. No big surprises here. It’s bigger tumors, more nodes, grade, lack of ER, those kinds of things. DR LOVE: And I’ve heard clinicians express confusion about how to use the 70-gene assay because they don’t have Adjuvant! Online. And granted, yes, there’s the appendix that you mentioned. But I’m not sure it really fundamentally changes the message here. DR BURSTEIN: I don’t think it changes the message. I think it does make it a little harder to operationalize this, because a tumor that is Grade 2 would be low risk in many reads but in Grade 3 would be high risk. And many people don’t exactly have so much confidence in grade that that feels comfortable. So what I’ve tried to do here is to break out the data by the clinical and genomic risk types. And those are labeled up at the top, clinical low versus high and genomic low versus high. And when you did the data like this, it made a few pretty apparent points. The first is that clinical low risk, as you’d expect, was predominantly node-negative, Stage I cancers, very few higher-grade tumors. And by contrast, clinical higher risk would be nodal involvement, typically higher-grade, larger tumors. And I think that’s, on the one hand, perfectly fair. On the other hand, as we’ve been talking about already today, when you have a little bit of disease in the nodes, it’s probably not the same as having a lot of disease in the nodes. The second point is that most of the data were from so-called luminal or ER-positive breast cancers, rather little experience here with triple-negative or HER2-positive tumors. And the vast majority of those were assigned to the chemotherapy arm. So it really still remains a study about ER-positive, HER2-negative cancer, which is an area that’s been well mined for the role of genomic assay. So let’s focus first on the group that had clinical low scores and a low genomic profile. Again, the majority of these were Stage I, ER-positive, Grade 1 or 2 cancers, and they did very well with endocrine therapy alone; 98% 5-year distant disease-free survival, very consistent with that West German PlanB study, very consistent with the TAILORx trial, a nice validation of another genomic assay as a prognostic marker. If you look at the triple-negatives, very little experience. They almost all were assigned to chemo. I don’t think the data speak to them, and I wouldn’t order the test there. The same is true for the HER2-positive subgroup. In the group of women who had clinically high stage but genomic lower stage, they actually, again, had a very good prognosis, 94 distant cancer-free survival at 5 years. The p-value here for benefit for chemotherapy was potentially statistically significant. Whether it was clinically significant or not is a question, but chemo still made about a 2% difference here. So you couldn’t exactly rule out a role for chemo. And the same is true for the other group of patients who had the clinical low risk but a genomic high index. So the question is, on the one hand I think you can look at this and say, “This confirms that overall, integrating a genomic signature allows you to figure out which cohorts of ER-positive disease have a good prognosis.” And that does include, in this experience, nearly half of the women who had nodal involvement. At the same time, it doesn’t immediately prove that they get zero benefit for chemotherapy, which has been one of the touchstones for the Recurrence Score in particular and allowed it to become so popular. So if you look at these outcomes in the chemo group and you look at the discordant groups — clinical high, genomic low or vice versa — the interesting thing is, there was no major clinical benefit, but you couldn’t prove there was zero benefit. And so I think that it still leaves open to longer follow-up whether or not there would be some late emergent signal that chemotherapy might have a role. A couple of other nuggets along the way: One is that in this interesting group of patients who had Grade 3 tumors but had a low genomic score, the outcomes were really good. And that’s also been seen in data with the Oncotype DX Recurrence Score at the population level, where local grade assessment of Grade 3 in combination with a low genomic score, usually the genomic score is more powerful. And I think that’s an example where you probably want to integrate the genomic test with your local pathology, because you might be a little bit misled if you just gave everybody who had Grade 3 cancers chemotherapy. So my takeaways from the MINDACT data, which are shown here in the Kaplan-Meier format, are that if you integrate a clinical assessment with a genomic assessment, you’re going to be able to make a better decision than with either alone, that in general, tumors that have low-grade genomic features are going to have a good prognosis whether or not they get chemotherapy. And so I think it’s a validation for the whole concept that we’ve been putting forward for the past decade now, that for ER-positive, HER2-positive cancers, you’re going to want both traditional stage information and a genomic platform to make the best decision for patients. DR LOVE: Any thoughts about this issue about the fact that there does seem to be a little bit of benefit with chemotherapy with the patients who had the 70-gene, I guess, low score, and whether — you hear people talking about these are all the same, but maybe there are subtle differences that would allow one to be predictive and another not to be. DR KING: Right. It’s a very good point. When MammaPrint was first presented and published, it was not advertised or marketed as a tool that allowed you to predict response to therapy. It was simply just for prognosis. And that was a limitation. We all understood that limitation. And we didn’t use it widely in our practice, and we still don’t use it widely in our practice in the United States, although they do use it quite a bit in Europe. And so whether or not they’ve been able to take that forward now to show prediction of response to therapy I think is a little bit hard to fully understand, simply because in this study, as Hal pointed out, they did include a small number of triple-negatives and a small number of HER2-positives. And we know that those 2 disease subtypes are going to tend to be more responsive. And so if you pulled those out of here and reran the analysis, I don't know if those numbers would still be there.
Correlation between 21-gene RS and outcome in large population-based analyses
DR BURSTEIN: I actually think that on the one hand, these studies have just mostly reinforced what we already thought we knew. But on the other hand, when you begin to think that this is from populations of people, it really is quite extraordinary. You always wonder, “Do the innovations that researchers make really translate into better decisions and better outcomes in the real world?” And the answer is, they do. So there have been a couple of sets here. This is an analysis from the US SEER population, where they looked at outcome as a function of Recurrence Score based on SEER data. This is breast cancer-specific mortality, because that’s what’s in SEER and easiest to chart. And what’s amazing here is, if you look at the little tiny numbers that are below the Kaplan-Meier curve, we’re talking about, on this particular slide, close to 40,000 patients in the data registry. And what you can see is that in terms of survival, higher Recurrence Scores are less good for survival than low Recurrence Scores. And, in fact, the breast cancer-specific survival for a woman whose tumor was low, less than 18, was less than 1% at 5 years. I actually doubt very few of us would say to a patient, “You have a less than 1% chance of dying of breast cancer in the next 5 years,” but that really is the case. It’s an extraordinary population-based observation. And then these data get to that issue of extending it to node-positive and node-negative. It makes the same point that integrating stage with age with genomic information, again, really teases out the prognosis. And again, if you look at both node-negative and, in this case, node-positive patients, breast cancer-specific mortality at 5 years, very low for the low scores and appreciably higher for the high scores. Now, given the interest in extended therapy and the late recurrence risk, especially for the node-positives, you may say, “This 5-year landmark is not the best analysis,” but those are the data that are available. And then they’ve done the same thing in a large registry cohort that happened to be based in Israel. The numbers here are not nearly as big. We’re talking about something on the order of 2,000 patients. But again, at the population-based level, distant metastatic risk clearly correlates with Recurrence Score. So I like these data, as I said, because I think it takes the observations we have from academic centers and research participants, which are inherently biased — we know they’re biased towards better educated, more affluent, overall healthier patients who go to treatment, and certainly those who participate in clinical trials can be described that way compared to the general population — but now showing that even if you take away a lot of those things and you look at the implementation of the technology across the US, you’re seeing corroboration of these remarkably good outcomes. DR KING: What I found really interesting looking at those, the SEER data that you were showing, is that the data are really starting to dichotomize. The low and intermediates are doing — both doing very well. DR BURSTEIN: Yes. DR KING: And the highs up here, in both the node-negatives and the node-positives, it’s almost like we’re getting to 2 groups. And given the fact that these are population-based data — and so some of the intermediates would have gotten chemotherapy, some of them would not, based on how their physicians treated them — right? I wonder if there’s not something to be learned from how those 3 lines are becoming more like 2 lines. DR BURSTEIN: That's interesting. Pathologists have often said there’s no such thing as Grade 2, right? DR KING: Mm-hmm. DR BURSTEIN: It’s just that that’s a fudge in between. DR KING: Mm-hmm. DR BURSTEIN: And the truth is, there’s probably, again, just this continuous spectrum. But one of the challenges in clinical research right now is that we’ve kind of entered this world of 2 worlds of breast cancer. There’s this low-risk group, which is fortunately a lot of patients, because of screening mammography and things, where they’re going to do very well. And it’s hard to even come up with randomized trials for them, because the event rate is so low. And then there’s this higher group by stage, for the most part, Stage II and III. And that’s where most of the research is going on right now. So that’s where the PALLAS trial is. And that’s where other studies of adjuvant manipulations are. And we saw this again in the data for ovarian suppression, which was if you had a very low-risk tumor, ovarian suppression didn’t add to tamoxifen because the natural history was such a favorable overall prognosis. And the predictor for benefit of ovarian suppression was things like did they have enough risk that you thought they needed chemotherapy? And were there nodes involved? And so this separation of breast cancer into more worrisome and less worrisome is the driver right now for most of the clinical research going on in the field.
Reliability of genomic assays for patients with high-grade tumors
DR BURSTEIN: So one of the interesting things in the Recurrence Score literature, as well as in the MINDACT trial, is that there are these groups of women who have Grade 3 tumors but have low genomic signatures. Actually, the West German PlanB data had a reasonably large cohort of these patients. The MINDACT data did as well. And what I think we have on the next slide is that grade and genomic score still matter — no big surprise — but that the Grade 3s actually did pretty well when the Recurrence Score was very low. What you want to focus on here is looking at the poorly differentiated or high-grade tumors. And you’ll see that the ones that had a low score, shown in green, again had a very, very low 5-year risk of recurrence. So in both specific studies and in the population, Grade 3 with a low genomic score looks more like a low genomic score cancer than it does look like a Grade 3 cancer. And I mention that because the discussion we have all the time with our colleagues around the world is, do we need these genomic signatures, or can we just use grade, ER status, Ki-67 to answer the question? And I actually think that the honest answer is, you do better when you have both pathology and a genomic index. DR LOVE: But, I mean, what you just said kind of makes you think, “Do you need grade?” DR BURSTEIN: Grade is the most operator dependent and least reproducible of all of these. DR LOVE: I mean, your algorithm has you not doing a genomic assay if they’re low, and yet you just said that it’s not necessarily concordant. DR BURSTEIN: You’re right. We feel more confident with the lower grades than with the higher-grade calls. But I think that grade and, similarly, Ki-67, which are the historic readouts, are the things that are under threat by the genomic assays. And one of the interesting things is that when you really step back from all the genomic data, there’s no surprises to any of this. Everybody would have said that a low-grade, low-proliferative tumor does better than all the others. And it wasn’t until, though, we had the genomic assay that was an independent arbiter that oncologists really sat up straighter in their chair and said, “Mrs Jones, you don’t need chemotherapy.” And I think that without that genomic test, it’s still very hard to feel confident about that recommendation.
Case: A 72-year-old woman with a 2.6-cm ER/PR-positive, HER2-negative lobular carcinoma with 4 of 4 positive sentinel lymph nodes and a 21-gene RS of 11
DR BURSTEIN: So this was an older woman who at 72 had been diagnosed with a lobular carcinoma. She ended up with a mastectomy — again, many of these older patients may elect that — and sentinel node mapping, where she was found to have a just over 2 and a half-cm low-grade, ER and PR-positive, HER2-negative cancer. But she had metastatic cancer in 4 out of 4 nodes. And sometimes these lobulars, they do these immunohistochemical assays because you see that kind of shotgun distribution of the metastases in the node. And then she had an Oncotype DX Recurrence Score sent, and it comes back low at 11. So this is that good biology but riskier stage. And the interesting thing is that this is not a low-risk cancer even though it has a low genomic score. In some of the data that Mitch Dowsett and others have done, like the TransATAC analysis, this woman probably has a 50-plus-percent risk of recurrence with endocrine therapy over the next decade. DR LOVE: Incidentally, this is a Grade 1 at your place. DR KING: Mm-hmm. Yes. DR BURSTEIN: Then it’s for sure. That is the archetype of a Grade 1. DR LOVE: Yes, but I’m just saying it’s interesting that they read it as Grade 1 and she has 4 positive nodes. And yet her Oncotype is low. So kind of all comes together. Any comments about specifically lobular, Tari, and what we know about data, particularly — I think I’ve seen data with the Recurrence Score specifically in lobular. DR KING: There has been efforts to look at the subset of lobular tumors. And to my recollection — Hal can correct me if I’m wrong, but I think that it predicts just as well. I don’t think that there’s any deficiencies looking at the lobular subtype versus the ductal subtype. These cases are challenging, though, because again, this is a 72-year-old. She’s had a mastectomy. Again, from a surgical perspective we want to do less, if we can. But this is somebody with 4 positive nodes. And in my practice, I would have recommended an axillary node dissection, either — if these would have been identified intraoperatively, we would have performed it. Postoperatively, I would recommend that she go back simply because this does suggest that there may be a higher volume of disease in the axilla than what we’re comfortable leaving behind, relying on systemic therapy to control. With a 72-year-old with a Grade 1 cancer, we would be having the discussion of whether she needs PMRT for 4 positive nodes. And if it came down to a decision between if I dissect her axilla and there’s no additional disease, then she can avoid PMRT, versus if I just leave you with these 4 positive nodes not knowing if there’s anything else there, if that was going to tip my radiation oncologist towards giving PMRT. I would certainly want to do whatever I could do from a local therapy point of view to reduce the likelihood that she would receive PMRT. DR LOVE: And Hal, I think I did see a series, particularly in lobular. And even though there were, I think, few, if any, high Recurrence Scores, there were quite a few intermediate. And I’m curious, for example, in this situation, if her Recurrence Score had come back as intermediate, would you have given her chemotherapy? DR BURSTEIN: I think you could have begun a more serious discussion about that. I think in this case, given the low grade and the low Recurrence Score and, as it happens, the patient preferences, it’s a tough sell for chemotherapy. But Tari and her team, who have been involved in the sequencing of these lobular cancers, have really unmasked a lot of heterogeneity. And while most lobulars have a pretty good biology, a few don’t. And a few are kind of nasty cancers. And these are not just what the pathologists used to call the pleomorphic lobulars, which probably aren’t lobulars — my take on what your data have suggested — but there are the higher Grade 1s. And a few of them even have these funny HER2 status issues. And so we’re learning a lot more about the heterogeneity within lobular cancers. And I think that that’s the kind of information we’re going to be looking to in the years to come. DR LOVE: So I’m going to guess that this lady did not get chemotherapy, because you wouldn’t have ordered the Oncotype. I’m curious, though — DR BURSTEIN: That’s an important point, because we do see these cases sent in for referral, like, “The patient had 28 positive nodes, but I got an Oncotype, and it was 4.” It’s like, “Why did you get the Oncotype?” I mean, at some point if you’re not going to use the information, you don’t need to get the test. And that may come up at the other end of the spectrum too, which is on these little tiny cancers that people sometimes get the test for. DR LOVE: But I am curious in terms of what Tari was saying, what you did about the axilla. DR BURSTEIN: Yes. So in this case, I agree. I thought she needed additional local therapy. We weren’t going to do chemo. And at that point, I thought that she should just proceed to radiotherapy, that this is, again, something that we talk about all the time at tumor boards. But she didn’t have any bulky disease at presentation. She had these kind of occult fossae of cancer in her lymph node. And I’m sure that if we did a dissection, we’d find more of it. DR KING: This is where that multidisciplinary discussion becomes so important, because if my radiation oncologist were to say, “I’m going to recommend radiation whether you dissect her axilla or not,” then of course I would say, “Okay. There’s no reason to give her the double axillary treatment.” But if they were to say, “If you dissect her axilla and there’s nothing else there, then we won’t give her radiation,” then I would like to try to avoid that radiation if I can, because again, we’re not just talking about radiation to the axilla. We’re talking about radiation to the entire chest wall. I don't know if she had reconstruction or not, but certainly radiation has impacts on reconstruction. It’s 5 more weeks of daily therapy. Depending on her other health status or where she lives, that may be inconvenient. And we do see, unfortunately, some radiation-induced malignancies. And so I’m very cautious. If there’s an opportunity to avoid postmastectomy radiotherapy by doing a little bit more surgery, then I want to at least explore that option.
Case: A 73-year-old woman who received anastrozole for a 14 x 8 x 8-mm ER/PR-positive, HER2-negative invasive ductal carcinoma (IDC) after a 21-gene assay revealed a RS of 22
DR KING: So again, 73-year-old went in for a screening mammogram. She is a product of 3-D mammography now, so this area that showed up in the lower inner quadrant was a pickup — as we say, a “tomo” finding. Further workup demonstrated that there was actually a mass there, a relatively small mass, 6 o’clock position, 2 to 3 centimeters from the nipple. A biopsy demonstrated Grade 2, ER-positive, PR-positive but not quite as strong, HER2-negative cancer. We discussed her care, again in a multidisciplinary fashion, because we do see our patients actually together in real time. And the decision was made between myself and her medical oncologist that she was a patient who the oncologist would lean toward really only giving her hormonal therapy. And whether or not she had a lymph node involvement was not going to change his recommendation. And so again, based on the CALGB data, we know that women over the age of 70 who have early-stage, ER-positive breast cancer, whether or not they have axillary staging or not, they do not seem to have increased rates of axillary failure that results in any detriment in their overall outcomes. And so there is a group of women that we can potentially avoid axillary staging if it’s not going to change their treatment planning. She was one that we decided that we could do this on. So we simply did the wire-localized lumpectomy, a 0.9-cm Grade 2 cancer. Oncotype came back with a Recurrence Score of 22. And this was a discussion between the patient and the oncologist as to how they felt about that intermediate score. And she ultimately elected to have hormonal therapy alone. DR LOVE: So just out of curiosity, if she’d had a high Recurrence Score, would you have rethought your axillary decision? DR KING: So it’s a very good question. I think the Recurrence Score, as we remember, predicts for a risk of distant failure. Now, there is some correlation with prediction for local failure, but that’s in a smaller data set where that’s been published. And we don’t really use the Recurrence Score to think about, “Should I do more surgery?” We use the Recurrence Score to think about, “Should I give more systemic therapy?” So the decision, if the Recurrence Score was high and the oncologist would have leaned toward giving her chemotherapy, I’m not sure that that would have meant that I needed to go back to do her axilla, because again, the treatment decision was already made. The question comes down to, what do I think her risk of axillary failure is or regional recurrence is? And with the Grade 2 small cancer, if, in the setting of breast conservation we know that again, 1 or 2 positive nodes, we’re not dissecting the axilla. We’re not seeing high rates of axillary failure in those women. And I would have felt comfortable observing her axilla, again, knowing that she’s getting the appropriate systemic therapy. She also would get whole breast radiation therapy, presumably, if she had a high score and that was leaning people towards more therapy. And so perhaps my radiation oncologist would have added a high tangent or added a radiation field if I did not go back and dissect her axilla. But these are the nuances that we get into where we have these multidisciplinary discussions about what really does the Recurrence Score mean? And I really do — I go back to Recurrence Score. A high Recurrence Score doesn’t mean a patient should have, for example, a mastectomy. A high Recurrence Score means the patient should have appropriate systemic therapy. And it’s the interaction between our local treatment and our systemic treatment and the biology of the tumor that really determines these long-term outcomes. DR LOVE: So before I get Hal’s thoughts on this, 1 more question about her, which is, would you have approached her differently if she wanted to have a mastectomy? DR KING: So if she would have wanted to have a mastectomy — that’s a very good question. It’s a little — we were right there and can feel her axilla. But it’s potentially yes, because again, I would have not wanted to put her in a situation where she may be looking at having radiotherapy because I did not take care of the axilla. So the CALGB data includes women who had lumpectomy with or without radiotherapy. That’s not a mastectomy cohort. In the mastectomy setting, I probably would have done the sentinel node procedure and sent it for frozen section, so that I would feel comfortable with her local management. DR LOVE: So Hal, any comments about this case and the algorithm behind it? DR BURSTEIN: So there’s a very nice analysis that the NSABP published after the B-20 data, making the point that not all Recurrence Scores are the same. So a 22 is not always the same result. If you have a smaller cancer with a lower grade, a 22 probably is an overestimate of the risk. And, by contrast, if you have a higher-grade tumor that’s larger, a score of 22 and the numbers that they put on those sheets probably underestimate the risk. And so you still need to integrate your traditional pathobiological information about stage and grade. In this case, for a woman who had a subcentimeter cancer, I think the thing that the Oncotype result here says is there’s probably not a huge benefit for chemo. We don’t really know what that nodal status is. But you’ve probably excluded a big benefit for chemo. And my guess is that was more than enough information here to make them very comfortable with endocrine therapy alone.
Use of genomic assays in the neoadjuvant setting for patients with ER-positive, HER2-negative BC; other clinical characteristics shaping neoadjuvant therapy decision-making
DR LOVE: The first question we asked people was, have you, in the last year, utilized a genomic assay in the neoadjuvant setting? And you can see that almost half of the medical oncology investigators had done it. Again, primarily the 21-gene Recurrence Score. And again, about half of the surgeons had used it, mainly 21-gene Recurrence Score. When we asked also the question about the incorporation of grade, Ki-67 and pathology in the neoadjuvant setting, Tari, this is what we found. I think it looks pretty similar to what we saw in the adjuvant setting. But any comments on that? DR KING: Mm-hmm. So I do think this is a very interesting question. And I was also intrigued by the results, because it’s something that, intuitively it makes sense, but I have to confess that I don’t think we do this very often, Hal. And I know when I was answering the questions I checked the box that I did it as part of a clinical trial, because we do have a couple of trials of neoadjuvant endocrine therapy where certainly, if you got a high score, you would not want to put that patient on the neoadjuvant endocrine therapy trial. But I think it’s a very compelling argument. And I’m actually happy to see that people are thinking about it like this. DR LOVE: Any thoughts? I don't know that you have been doing this. DR BURSTEIN: We don’t get it very often, as Tari just alluded to, outside of clinical studies. And we certainly see a lot of consultations where this comes in. I’m not sure that the testing here helps the clinical decision-making. And so on the one hand, I really think that the lessons from genomic testing apply across the whole spectrum of disease from preoperative to adjuvant and, as we’ve said, even to metastatic and probably a lot of things in between. Having said that, the issue for neoadjuvant therapy is, you’re usually talking about women who have higher-stage tumors and where you need a clinical result to shift their surgical treatment options. And that’s where I think some of the more provocative data are just coming forward from Harry’s study and from some other trials that are going on. So I think we’re going to be seeing more of this, though at the moment I’m not sure how much hard data there really is to tell us. DR LOVE: But it kind of, Tari, gets into the biology versus anatomy thing, because I always thought it was kind of weird that here’s this tumor, and you’re trying to decide, ER-positive, HER2-negative, maybe you’re even thinking about neoadjuvant hormonal therapy. And so, okay, we don’t get genomic assays. We make the decision. Then let’s say the patient goes to surgery. The same exact tumor, we use the assay to decide what the biology is. DR KING: Right. DR LOVE: It seems like maybe we want to know the biology preop. DR KING: Yes. I agree. I think maybe I’m more enthusiastic about it, because from a local therapy standpoint, we are seeing such a shift in our approach. So patients with HER2-positive disease and triple-negative disease, the majority of them we are treating preoperatively because we know they’re going to get therapy anyways. And so why not take advantage and potentially get that secondary benefit where it may alter our local management? And in the ER-positive setting, we have not applied that. And we know that patients are less likely to achieve path CRs, but some of them still will. And so if we know preoperatively that the patient’s going to get chemotherapy anyways — and many times we just have nothing to lose by trying it preoperatively and seeing if we can get that secondary benefit. Now, I’m not saying that everybody with a 1-cm tumor needs to consider preoperative therapy, but certainly in the higher disease stages, if you have an ER-positive patient who you know is going to get it, I would like to see if we can get some secondary gain from using it up front.
Role of neoadjuvant therapy in improving surgical options for patients with BC
DR BURSTEIN: So in B-18, the classic study of neoadjuvant therapy, we learned 3 things about neoadjuvant treatment. It was a study comparing surgery first versus AC chemotherapy back in the day versus AC and then surgery. We learned 3 things: (1), that the long-term results were exactly the same. So that enabled all these trials of preoperative therapy, because you weren’t hurting anybody. It wasn’t better. The original Fisher hypothesis was that up-front chemo might actually be better, because you would mop up microscopic bits and all that sort of thing sooner. Didn’t prove to be the case. The second was that we learned that you would find a surrogate for prognosis. So the better the response — and, in particular, the complete eradication of the tumor — the better the long-term prognosis. And the third point was that you could surgically downstage many patients. We’ve learned nothing since then in terms of outcomes for neoadjuvant therapy. We know the exact same amount now that we do then. We’re willing to use it more. And there’s a lot of literature about it. And there are even novel drugs approved for this indication. But in terms of the fundamentals, that’s all been there for 20-plus years. And the primary reason to use neoadjuvant therapy in 2017 is to improve the surgical options for the patient. And increasingly, because of all the imaging we’re doing and the patient selection, it’s even less about mastectomy versus lumpectomy than it is about preserving the axilla, because for clinically node-positive patients, the opportunity to surgically downstage them and spare an axillary dissection has been the greatest appeal, I think, of neoadjuvant treatment. DR LOVE: Maybe you can just briefly comment on that, which is, I guess, there are 2 critical scenarios. One would be the patient, as Hal was just suggesting, with clinically positive axilla, biopsy-proven axilla. You give them neoadjuvant therapy. They have a clinical CR. It’s a path CR. The sentinel node postneoadjuvant therapy, which now seems to be part of practice, is negative. The patient who had biopsy-proven axillary node is avoiding an axillary dissection. Any comment about that scenario? DR KING: Yes. So it really has been a paradigm shift to take someone starting out with a clinically positive axilla and give them neoadjuvant therapy and then to even perform the sentinel node and trust the result. And I think that’s where as surgeons we have to be very careful, is, how confident are we that that sentinel node reflects the status of the axilla? And there have been many efforts to demonstrate technique issues to make sure that you’re evaluating the axilla and that your false-negative rate is going to be low. And if you apply those, yes, it has been adopted. And people are avoiding axillary dissection. The next step is, does that patient still need radiotherapy or not, because traditionally patients who start out with positive nodes and received neoadjuvant therapy, they would have also been facing a decision regarding radiotherapy, whether it’s PMRT with nodal radiation or whole breast radiotherapy with nodal radiation. And so we have a very important randomized trial going on right now through the NSABP and the NRG, taking exactly that patient who achieves a path CR in the nodes and then randomizing them to radiation to the nodal fields or not. And that’s a very important study, because we’d all like to dial back. But as Hal said, we don’t have a lot of outcome data to give us that reassurance that we can dial back. And B-18 and B-27, the long-term follow-up from those studies are still really the cornerstone of how we make those treatment decisions for radiotherapy after neoadjuvant therapy. And we know from those studies that someone who remains node-positive after neoadjuvant therapy has a very high risk of local-regional recurrence, whereas someone who we’re clear is node-negative, that risk of local-regional recurrence falls down below 10%. And so really going back to the surgeon and the team and being very important that not only are you proving that someone did have a path CR and that the nodes are negative but also making sure that you don’t miss the patient that still has a positive node after neoadjuvant therapy, because that has significant impact on their subsequent treatment decisions. DR LOVE: So I want to get into a little bit more in terms of the decision about neoadjuvant therapy. But one other local therapy issue I was curious about is, Tari mentioned the issue — of course we went through the axillary issue, but also the issue of the primary breast surgery and neoadjuvant therapy. And patients who, quote, are even candidates for breast-conserving surgery, the potential advantage in terms of margin and the type of surgery. Any comments about that? DR KING: Yes. So it is a conundrum that as surgeons we face. We see patients who are perfectly good candidates for a lumpectomy, sometimes with or without neoadjuvant therapy. And patients have been leaning towards more surgical treatment. And so we’re seeing higher rates of mastectomy, higher rates of bilateral mastectomy. And we’re really trying to tease out and understand what’s driving those decisions. But in the neoadjuvant setting, the thing that now is coming into play is that we are seeing these increasing rates of path CR, particularly not in the ER-positives but in the HER2-positives and the triple-negatives. We’re seeing 40% to 50%, sometimes 60% of women with no residual disease in the breast, and yet many of these women are still choosing to pursue a mastectomy. I think the next phase, the next set of really interesting and provocative trials will be, can we prove that someone has a path CR in the breast and potentially avoid breast surgery altogether? And there are some, again, studies coming through some groups looking at randomizing patients to whole breast radiation instead of surgery, because it really would be nice to take advantage of these high path CR rates that we’re seeing in the breast and, again, reduce some of the morbidity of our surgical procedures.
Neoadjuvant therapy recommendations for a younger patient with ER-positive, HER2-negative BC
DR LOVE: So I want to explore a little bit some of the questions we tried to look at here. So here you see the 2 left sides, you see node positivity, 1 with a palpable lesion, 1 by imaging and then clinically node-negative with different tumor sizes. So I think what we see is, first of all, in general in the medical oncology investigators a fair amount of use of — quite a bit of use, actually, particularly in larger tumors, of just chemotherapy. Others just sending the patient to surgery. But again, not an insignificant number of people using genomic assays in the node-positive situation, particularly with the smaller node-positive tumors. In the clinically node-negative, you see a lot more use of genomic assays and up to almost half of the investigators using a genomic assay or just sending the patient to surgery, maybe not quite so much in the surgeons. Hal, any thoughts, first of all, about what we’re seeing here in the 40-year-old patient? DR BURSTEIN: The key data set here is the — or the number is the 40, right? That’s the 40-year-old. Because even when we start talking about outcomes using genomic predictors in ER-positive, node-positive patients, most of that literature is in historically postmenopausal women. And again, we think the same principles will probably apply down the spectrum of age. But to the extent that chemo has a role, it’s more prominent in younger women, perhaps because it actually induces menopause, perhaps because there’s a little bit of different biology in these women on average. But a 40-year-old with a clinically palpable node, for most people the standard of care is going to be to think about chemotherapy. And you really have to want to twist yourself around into an unusual case situation before you really were taking that off the table. So I confess I don’t get a lot of Recurrence Scores in this setting if the node is obviously positive, because most of those women, I think, warrant chemotherapy. DR LOVE: What about in the patient with clinically negative axilla, Hal? DR BURSTEIN: Again, for the most part we’re still making this decision based on traditional clinical factors. Do they need a response up front, such that we want to improve their surgical option? If they’re clinically node-negative with a 2-cm cancer, most of those women just go to surgery. So this is a conversation. The other thing that happens in these younger women is that almost all of them seem to have had an MRI. We don’t actually order a lot of MRIs, but they all come with their MRIs. And on the MRI, there often are multifocal areas. There are other spots that people want to evaluate. A lot of these women feel like they’re going to end up with a mastectomy anyway, just because of the nature of their presentation. That doesn’t mean it’s wrong to give neoadjuvant chemo, but I think it plays into some of the dialogue about the decision-making and how much things maybe could change, but often don’t, based on a response.
Weighing the value of existing genomic assay results for a patient for whom a genomic assay might not be ordered
DR LOVE: So now we get into kind of a particularly interesting part of these data, because we started to see this at meetings, situations where we would present a scenario and say, “What do you do?” And one of the questions would be considering a genomic assay. And we would create scenarios where we knew most of the audience would say, “We don’t use a genomic assay.” And then we would say, “Oh yes, let’s go back. And now that’s the same patient, except they’d already had an assay,” and to see whether — it shouldn’t theoretically change what they do, I think. So we tried that here. So this is a node-positive patient, palpable lesion, 2.8 centimeters and, in this situation, most people would use — 62% would just go ahead and use neoadjuvant chemotherapy. Very few would order a genomic assay. We go back and we say, “Here’s the same situation, except the patient actually already had a genomic assay. And it was the Recurrence Score, and it was low.” And you see people change their decision. And I think we saw the same thing in the surgeons. So I don't know. Does this in some way, Tari, seem inconsistent? DR KING: Probably not inconsistent, but it’s not an area where we’ve traditionally used the assay. And so people feel like they can make their decisions again. And as a surgeon, when I think about neoadjuvant therapy, I’m thinking about what can I gain from it from a local perspective? I tell my patients that there’s no difference in terms of survival, whether you get chemotherapy before or after, if your oncologist thinks that you need it. And so let’s think about what’s the secondary gain here. And so again, in someone with a known positive node, if I have a genomic score that says that she has a low likelihood of being sensitive to chemotherapy, because really the low scores tell us that there’s really not a likely benefit, then that would probably sway me against telling her that I think we’re going to gain a lot from giving the preoperative therapy. Because I’m really looking for a dramatic response if I’m going to use preoperative therapy in someone who’s otherwise a candidate for lumpectomy or someone who has 1 positive node. So if I have information that tells me she’s not going to have a dramatic response, then I’m going to feel I guess less compelled to have her get that therapy preop. Also, I will point out that our management strategies for the axilla are very clear in the up-front surgery setting. So we have data from Z11. We have data from AMAROS that if someone only has 1 or 2 positive nodes, I don’t have to do additional axillary surgery. Once we muddy that water a little bit with the chemotherapy, then the decisions about what happens if there’s 1 or 2 positive nodes left over becomes a little bit less clear. So maybe not so much in the biopsy-proven palpable setting, but back on your previous slide where in the clinically node-negative setting the surgeons were more likely to just to go surgery. And I think that’s a reflection of we have really nice data in the clinically node-negative setting that helps us dial back on axillary treatment. Once we use neoadjuvant therapy in clinically node-positive, clinically node-negative, those decisions afterwards are not as clear. We don’t have as much data to support them. DR LOVE: But, I mean, you were one of the people who switched what they do. DR KING: Absolutely. DR LOVE: So when we asked, we said in general, yes, I would give chemo. And then we said the Recurrence Score is low. I mean, I guess you could argue maybe it’s not that likely, but it seems like it wouldn’t be rare. So how do you feel about your own decision-making, knowing that? DR KING: All right. As I said, if I’m going to give chemotherapy, or if I’m going to ask my medical oncology colleague — DR BURSTEIN: You can give it. As long as I don’t have to do any operations. DR KING: If I’m going to suggest that a patient with ER-positive disease get preoperative chemotherapy, it’s because I’m hoping that she’s going to have a dramatic response and that we’re going to achieve a path CR in the axilla, so that I can avoid axillary node dissection. So if I then am given a piece of information which tells me that she’s very likely to achieve a path CR in the axilla — and I think we can extrapolate from what we know about the Recurrence Score and prediction of response to therapy — if I have someone with a low Recurrence Score, I think there’s a very low likelihood that she’s going to achieve a path CR in the axilla. And, therefore, I may not feel as strongly about her getting that preop therapy for that potential secondary gain, because that potential secondary gain is now very, very unlikely to happen. DR LOVE: So just to be clear, though, in this situation, you still would not use a genomic assay, or are you rethinking it? DR KING: As I said, it has not been part of our practice. But I think it’s particularly intriguing, because again, my goal in the preoperative therapy is, what is the secondary gain from surgery? DR LOVE: So Hal, we actually saw a very similar thing with the same kind of scenario saying, “Suppose they had a 70-gene assay?” And you see a big change in what people do. But again, help me think this through. DR BURSTEIN: So I think we’re — it’s reassuring that they picked both assays and drew the same conclusions. But I want to get back to the point Tari made, which is, the real issue here is about surgical management. If you have a low score, it probably means that from a distant metastasis recurrence point of view, chemotherapy may not change things very much. But that does not mean 2 things. First of all, it doesn’t mean the tumor won’t respond to chemo, because tumors shrink in response to chemo even if they don’t completely disappear. We’ve known this for a long time. If you look at luminal A tumors, there’s less likely to be a complete pathologic response than if they’re luminal B. And there have been other reports over the years that low Recurrence Score is much less likely to be associated with complete pathologic response. Luca Gianni did this in a very nice study 10 years ago now. So there’s no doubt that low score predicts lack of complete pathologic response, but it doesn’t predict lack of clinical benefit. You can still shrink the tumor with chemo even if there’s some there. The change of complete pathologic response with endocrine therapy up front is zero. So if your goal is to achieve a path CR, what this tells you is, you’re not going to get there with either chemo or endocrine therapy. It’s very improbable. So then the question really isn’t about achieving path CR. That’s not the goal of neoadjuvant therapy. The goal of neoadjuvant therapy is to make your surgical colleagues’ job easier, so they can do less surgery. And what is hard to know in a case like this is, which is really better, chemo or endocrine therapy for achieving that? So that’s the issue. So if you say, “Okay. I did the genomic assay. It’s really low. I want to give them up-front endocrine therapy to shrink that node and make the cancer smaller and enable surgery,” great. But there’s still going to be residual disease. And in a very old literature, the optimal endocrine result was after a year of neoadjuvant endocrine therapy. Few of us feel comfortable waiting that long in a 40-year-old woman who has a palpable node. And we don’t really know that no chemotherapy is the standard for such patients yet. So I just think you have to be cautious about overinterpreting a genomic assay in this particular situation. And again, we often hear it said that ER-positive tumors with favorable genomic profiles don’t respond to chemo. That’s not true. The cancer will get smaller. It gets spongier. It gets smaller. The nodes resorb. There’s still going to be residual disease. That’s the big difference.
Practice patterns in the neoadjuvant setting for older patients with ER-positive, HER2-negative disease
DR LOVE: So just to let you take a look again at the numbers that we got with age — Here are the surgeons. We go up to age 75, and now you see a lot of red and not too many people thinking about neoadjuvant therapy. There was another angle that we had about this thing about the doc who’s not using a genomic assay yet would be affected by the result. And that’s the flip side, which is the older patient with the high Recurrence Score. So we present a patient, 75 years old, with a palpable biopsy-proven node. And most people would not use a genomic assay, but about a third would in this neoadjuvant situation. And yet when we said, “Okay. The result was high,” most people would use chemotherapy. So a lot of the docs who said they would just send the patient for surgery, now knowing the Recurrence Score is high would give neoadjuvant therapy. And we saw that in the surgeons. Any thoughts about this? DR KING: Again, this is somebody who if the score was obtained in the adjuvant setting, there would be a very serious conversation about using chemotherapy. And so if this information is available in the preoperative setting and I think that I can get a secondary benefit, which is potentially downstaging that axilla, achieving a path CR in that axilla and avoiding the axillary node dissection in the 75-year-old who’s already probably going to get chemotherapy, then I would discuss the potential secondary benefit, again, from using it preoperatively. But I still think that there’s room to go either way here. If you have the information, I think you should use it. Do I think we should routinely be getting this information in a 75-year-old? Probably not. DR BURSTEIN: But I think this is a more compelling case for the genomics than the previous one, because it’s a positive result in the sense it says chemo really might get you to what you want in terms of that complete pathologic response. But there’s no mystery to this. We’re in Miami, so people do these charter fishing things, right? And if you go to your favorite fishing spot and you throw down a line and nothing bites, after a little while you move on. Or if you do the Loran thing and you know there are fish down there, you’re going to push a little harder. That’s all we’re doing here. What you’ve gotten in this result is a test that says, “At 33, that tumor’s not, at least historically, so likely to respond well to endocrine therapy,” and so chemotherapy really might be the much more powerful tool here. And you would never have given it to a 75 — not never, but you would rarely have given it to a 75-year-old otherwise. So you’ve found a piece of information that I think really does inform a better treatment decision here, though it’s pretty rare to see a 75-year-old with a score of 33. DR LOVE: So again, you fall into the category of one of these people who would not have used neoadjuvant chemotherapy, would not have utilized a genomic assay, and yet if you knew the result of the 21-gene Recurrence Score was high, would have used chemo. You’ve changed what you would do. DR BURSTEIN: Yes, but, I mean, that doesn’t bother me. If the patient presents with a breast mass and an axillary node and I’m talking about neoadjuvant chemo, we get the pathology back and it’s lymphoma, I’m going to use chemotherapy there, too. And this is a different beast. DR LOVE: But my question is, would you still, in this situation, not use a genomic assay? DR BURSTEIN: Oh. I think many of us might have been tempted to try endocrine therapy here if we didn’t have the genomic test. Sure, because in a 75-year-old with intermediate-grade cancer, you would have said, “Oh. There’s a very good chance that endocrine therapy will work and we can avoid chemotherapy if we get a clinical response.” Truthfully, I think a lot of these women just would have gone for surgery and an axillary dissection, if you actually were to do a real survey of what people actually have happen to 75 years — DR LOVE: But again, right now in this scenario, in general, you would send the patient to surgery? DR BURSTEIN: This is actually a place where actually we often use neoadjuvant endocrine therapy. I would have said this — neoadjuvant endocrine therapy is much more discussed than actually delivered in general. But older women with positive nodes, locally advanced breast cancer, that’s a place where we actually do frequently use it.
Results of a prospective multicenter trial using the 21-gene RS assay to choose neoadjuvant therapy for patients with hormone receptor-positive, HER2-negative tumors not suitable for breast-conserving surgery
DR BURSTEIN: So what they did is, they took a cohort of something on the order of about 50 patients. And they assigned neoadjuvant treatment based on the Oncotype DX Recurrence Score. All these cases were of ER-positive, HER2-negative breast cancer. If the score was low, less than 11, the women got neoadjuvant endocrine therapy. If it was high, greater than 25, they got neoadjuvant chemotherapy. And if it was in between, they were randomized to get endocrine therapy alone or chemotherapy. And then they charted clinical response, complete pathological response and the rate of breast-conserving surgery, which is probably the most important observation that one could make of such a cohort. DR LOVE: So it kind of looks like the TAILORx study, but neoadjuvantly. DR BURSTEIN: It’s a neo-TAILORx, though it wasn’t officially sponsored by the same group. And they didn’t include that in there. Now, before launching into this, you also have to acknowledge there’s a tremendous selection bias into this cohort. So these were patients where, for whatever reason, the investigators thought they would be appropriate candidates for endocrine therapy up front. This may not be a very typical patient population. But the intriguing thing is that the results are, on the one hand, very predictable but also really interesting. So if the score was low, less than 11, actually 85% of them had a clinical response to endocrine therapy. Zero had a complete pathologic response. And the rate of breast conservation was 75%. Now, we don’t know what that rate would have been if you just took them to the OR on Thursday, but — DR KING: They make a comment in their eligibility criteria that all women were considered not to be eligible for breast conservation but that that was based on size of greater than 2 centimeters. DR BURSTEIN: Yes. DR KING: So it’s really hard to know how — there are plenty of people with 2- or 3-cm tumors who are candidates for conservation, but they do say that, that they were felt to be not suitable for breast conservation based on tumor size. The one thing they don’t tell us in here, though, is the clinical nodal status or the pathologic nodal status pretreatment. We don’t have any information about that. So when they tell us that there were no patients that achieved a path CR in the breast and the nodes, it’s hard to know what that means, because I don’t know how many started out — DR BURSTEIN: Yes. No. You’re right. And a lot of these studies that have asked the surgeons beforehand, “Would you do breast-conserving surgery now,” and they would say, “No,” and then afterwards — it’s really a test of their willingness to do breast-conserving surgery more than the intervention. But three quarters were now able to be candidates for breast-conserving surgery. If you looked at the other extreme with neoadjuvant chemo, we’re talking about 14 patients. So you don’t want to overgeneralize this. But again, very high clinical response rate of 90%. Actually, a path CR rate of about 15%, very consistent with the literature on ER-positive tumors. But only 57% were candidates for breast-conserving surgery. Why that’s different is hard to know. So then focus on the group in the intermediate range. Of the women who got hormonal therapy, there were 18 patients here. About 50% had a clinical response. And 70% to 75% got breast conservation. Of the women who got chemo, actually there was a greater likelihood of response. About 75% had a treatment response. But fewer actually had breast conservation. So what do you take away from this, and how does it line up with other things we thought we knew? So the first is, on the one hand, it’s very predictable for the outcomes with chemo. We know that neoadjuvant chemo rarely achieves pathologic complete response, especially in low Oncotype DX scores. And you brought along this very nice paper from Denise Yardley and her colleagues, done a few years ago, where they showed the same thing. They looked at PCRRR as — Recurrence Scores a predictor of PCR in women getting chemotherapy with ixabepilone and cyclophosphamide. And what they showed here is that if the Recurrence Score was low, essentially you never saw complete pathologic response. And it was only when the score was high — in this case, greater than 30 — that you began to see even a 15% to 20% rate of complete pathologic response. Where are we going with all this? A couple of points: Chemotherapy rarely achieves PCR in women who have low-end Oncotype DX scores. If you’re thinking about getting that PCR in the node, you’re probably not going to get it with chemo. You’re not going to get it with endocrine therapy, either. But what you can get is effective tumor control. And the opportunity to increase the rate, perhaps, of breast conservation in women who had these very low scores and got endocrine therapy is the intriguing part of this to me. So I’ll pause there, and Tari can tell us her take on this. DR KING: I agree with everything. I think that this paper by Yardley, I think one of the other interesting observations is — and this is something that we’re seeing in multiple data sets — is that patients who have residual disease after neoadjuvant therapy are actually more likely to have residual disease in the breast, and not in the nodes. Now, some of these data sets are confounded by the fact that we don’t know who all started out with node-positive disease. But you can see here in this table that overall, residual disease in the lymph node with a negative breast was 9%. But residual disease in the breast with negative lymph nodes was 30%. And again, seeing this in other data sets where even though we don’t always achieve path CR in both the breast and the axilla with neoadjuvant therapy in ER-positive disease, there still is a significant minority of patients — DR BURSTEIN: That downstages the axilla. DR KING: — that we can downstage the axilla. And I think we have to pay more and more attention to that as more and more data sets come out and more people examine their own experience, because that’s really where we’re seeing the biggest impact of a change in our local therapy. All women want to avoid axillary dissection, risk or lymphedema. And so I just think that’s really something that we haven’t focused on that we need to start focusing on now.
Case: A 57-year-old woman with a history of multiple sclerosis who presents with a 2-cm ER/PR-positive, HER2-negative IDC and a 1-cm palpable axillary node
DR BURSTEIN: So this patient has MS, which makes the decision-making much more complicated. And you and I have talked about this over the years about curbsides, but we get emails all the time with curbside questions from people around the world. And what’s interesting to me is that most people actually get the basic things really well. So the curbside questions are always a curveball that takes you. It’s like it’s a perfectly normal case, but the Oncotype is 2. Or it’s a perfectly normal case, or the patient has multiple sclerosis or some other ailment. So this was one of those discussions that we had with the patient in consultation. So she had a breast mass, 2 centimeters approximately, Grade 2 on core, pathology ER-positive, PR low-positive, HER2-negative. She had a palpable node at presentation and FNA-confirmed cancer in the node. And she is pretty functional, but is on the bubble, right? She can drive herself, but she doesn’t drive at night. She uses a cane to get around, but she can still get around. And so here, there’s real concern that chemotherapy would be the tipping point to take a vigorous 57-year-old coping well with MS or some other illness and push them to a place where they really couldn’t. So in that instance, you’re looking for a real reason to use chemotherapy. And so here is a situation where I think an Oncotype could be helpful, because if it’s a low score, you’re making the argument to spare the patient chemotherapy, which would be a big win in the clinical situation. DR LOVE: So she had a score of 20. What happened? DR BURSTEIN: She began neoadjuvant endocrine therapy. And she’s had a partial kind of response, and there’s still some tumor there. The node’s a little smaller. And she’s done fine, obviously, with the treatment. I think she’s going to end up with a mastectomy and some kind of an axillary dissection regardless, but at least we have spared her the potential functional quality-of-life setback with the chemotherapy. DR LOVE: She got an AI? DR BURSTEIN: She’s on an aromatase inhibitor. DR LOVE: So Tari, any comments about this case and the issue of patients with comorbidities, the very elderly? DR KING: Yes. So it is challenging. Again, you’d like to avoid extensive surgical therapy. You’d like to minimize her morbidity. She’s walking with a cane. I don’t know if this was a right or left, if this was her dominant side, but you’d certainly like to minimize the amount of axillary surgery, particularly if this was on her dominant side. But as Hal said, chemotherapy is very challenging. So I would not push chemotherapy to try to get a path CR in the nodes, but I would be thoughtful about my approach to the axilla and discuss all the potential outcomes with the patient with the different scenarios. DR BURSTEIN: And here’s a case, just thinking down the road, where even assuming she has residual disease in the nodes, you might do radiation instead of a completion dissection, at least I’m thinking that here. DR KING: If she were to have breast conservation, then one could certainly consider that. I mean, it is a biopsy-proven node up front, so again, does not fit Z11, does not fit AMAROS. But if you evaluate her axilla and she has low-volume disease, potentially you could do a low Level 1 axillary dissection and, if there were any additional disease, then potentially using radiation therapy. But certainly I would want to minimize any functional impact to her arm.
Case: A 49-year-old woman with a 2.5-cm ER/PR-positive, HER2-negative IDC who underwent neoadjuvant chemotherapy after receiving a 21-gene RS of 29
DR KING: Also a 49-year-old. She is still premenopausal. Palpated a mass in her lower inner left breast. This was actually quite low, down near the inframammary fold and was not actually seen on mammogram. But on ultrasound you could see the about 2 and a half-cm mass there, biopsied, Grade 2 IDC, ER/PR-positive, HER2-negative. She had been sent for an MRI prior to seeing us. The MRI also demonstrated the disease, suggested it was potentially a little bit bigger with some enhancement of the underlying pectoralis muscle, again in the lower inner quadrant. She’s not a large woman, so there’s just not much tissue there. So I think it was just guilty by association, because on physical exam, the mass was clearly mobile. It did not feel affixed to the chest wall at all. But it was difficult to truly appreciate the size. She was, I think, a borderline candidate for one of our preoperative endocrine therapy trials. And so the decision was made with her medical oncologist to send the Oncotype, because if that Oncotype was high, then we would clearly be leaning towards preoperative chemotherapy. If that was low, then she would be a good candidate for this endocrine therapy trial. And so, in fact, her Oncotype came back as a score of 29, and she’s now on preoperative therapy. DR LOVE: So Hal, any comments about the case? Would you have used a genomic assay in this situation? DR BURSTEIN: I mean, certainly as it shook out it’s helpful, right, because you feel better about using the chemotherapy here. In a premenopausal woman with a 3-cm cancer and there’s this MRI question of adherence to the — I mean, it certainly would have been reasonable to just give the chemo also as the default. The clinical trial Tari alludes to is a very interesting study, but it’s predicated on the idea that you’re going to have a favorable prognosis, lower genomic index cancer. DR LOVE: What is the trial? DR KING: It’s Otto’s trial, the KELOPS, yes. DR BURSTEIN: It’s a risk stratification study where you get endocrine therapy up front if the tumor has a lower Recurrence Score.
Case: A 63-year-old woman with residual disease at the time of surgery after receiving letrozole for a 3-cm ER/PR-positive, HER2-negative, node-negative BC
DR BURSTEIN: So here we have a 3-cm node-negative presentation, higher grade, ER-positive, PR-positive. Ki-67 was tested and came back at 15%, which is an intermediate number. The question has been, would she need a mastectomy or neoadjuvant chemo? She wants to avoid chemo. And so the Oncotype was sent, looking to see if this would be a good setup. It came back at 20. She received 4 months of neoadjuvant letrozole and the tumor responded, and then she had a lumpectomy. But at the time of surgery, there was still residual disease, 2 centimeters of cancer in the breast, still higher grade, ER-positive and PR-negative. Loss of PR is something we frequently see in treatment of patients with aromatase inhibitors, because PR expression is dependent on functional ER signaling. And so the question that arises frequently in these kinds of cases is, okay. We had neoadjuvant endocrine therapy. There’s still residual disease. Now do you want to give them chemo? And this is a question that we often kind of agonize over. To go back to the framework we’ve put in place, if you had a 2- to 3-cm cancer and the Oncotype DX score was 20, you would have said she doesn’t need chemotherapy. Now you’ve done the additional experiment where you’ve offered them 4 months of treatment, and you’ve kind of had a response, but you’re wondering, “Is that as good as we can do? Would extra chemotherapy top off this patient in a way that’s clinically productive or not?” So we see this, these kinds of discussions when you invoke neoadjuvant endocrine therapy. DR LOVE: So before telling us what you actually did or are going to do, just to go back, though, it sounds like the surgeon wanted to see some tumor shrinkage, thought it would improve the cosmesis. DR BURSTEIN: Mm-hmm. DR LOVE: You said, quote, there was a little bit of a response. Do you think it actually changed and it allowed breast-conserving surgery or assisted the surgeon? DR BURSTEIN: I think it enabled it in the sense that they were looking for a reason to do less. They got something, and they were able to do less. DR LOVE: Any comments about this decision and the effect of neoadjuvant therapy and breast conservation? DR KING: So it really comes back to, again, a combination of what we technically can do and what is the patient’s preference. And if a patient is particularly motivated to try to preserve the breast, then we want to have those discussions and see what our options are. Do we need shrinkage of the tumor? Sometimes patients are just borderline candidates. And you really maybe just need a little bit to get her there, to feel like she’s a better candidate for conservation. And so when you have a patient who’s motivated to pursue that, you don’t have — again, I don’t think you have anything to lose by trying preop therapy and potentially extra or secondary benefit to gain. DR LOVE: Any comments about the issue of postneoadjuvant systemic therapy for residual disease, Hal? We have the CREATE-X trial out there that used capecitabine. How are you approaching these decisions right now? And how does it vary based on the ER/HER2 subtype? DR BURSTEIN: So part of it was — that’s what this case was, a teaser. DR LOVE: Right. I’m curious what you did there. This is a little different, because she hasn’t had anthracycline/taxane chemo yet. DR BURSTEIN: Correct. Yes. So for patients who get neoadjuvant chemotherapy for ER-positive, HER2-negative disease, if they have residual disease we tend to say, “Don’t worry. You’re still getting endocrine therapy. The most important piece is yet to come.” For women who have triple-negative disease, the issue looms a little larger, both because you worry more and you don’t have anything standard to offer. And occasionally we have thought about the capecitabine use based on the CREATE-X data. That data remain unpublished to date. It’s a one-off kind of study that was positive. It flies in the face of what most people were guessing capecitabine might actually do, though there was a very interesting poster at San Antonio, that a meta-analysis might suggest there is a role for capecitabine in triple-negative. But it’s a — I wouldn’t say a free-for-all, but it’s an uncharted area right now. Our standard approach is not to give more beyond standard anthracycline and taxane-based chemotherapy in those patients. DR LOVE: And I guess there are trials now looking at this. I think there’s a big one now comparing platinum agent to capecitabine in triple-negative. DR BURSTEIN: Correct. ECOG or SWOG. I forget. One of the cooperative groups is doing that. And then there’s also going to be looks at the immunotherapies in those cancer patients. DR LOVE: Right. I think there’s a pembrolizumab versus control. DR BURSTEIN: Right. DR LOVE: But in the HER2-positive space, you also have T-DM1 being looked at. DR BURSTEIN: Yes. I don't know if they’ve completed that study or if it’s — DR KING: It’s closed. DR LOVE: I know it’s trastuzumab — the KATHERINE trial. DR BURSTEIN: Right. The good news is that for HER2-positive tumors, we’re now seeing complete pathologic response rates of 60% to 80% in routine practice. So it’s less of an issue, because the up-front therapies are so effective. We still discuss, for instance, if a patient has had pertuzumab whether that should continue into the adjuvant setting, and there is some variation in practice. But we don’t know what more therapy would do at this juncture. DR LOVE: Any comments about this issue of systemic therapy for residual disease after neoadjuvant therapy? DR KING: Oh, I think it’s a very important area. And you mentioned a couple of the larger randomized trials. There are also several efforts at single institutions to really study that residual disease, and then you triage patients into one of several maybe earlier phase studies. I know MD Anderson is doing this. Memorial was starting to do this. So really understanding the biology of that residual disease is critical. And we certainly know that it has a bigger impact, as Hal said, in the triple-negatives, having residual disease much more significant than in an ER-positive having residual disease, where we know we have this endocrine therapy to use still.
Systemic therapy decision-making for patients with a local recurrence of BC
DR BURSTEIN: It’s a very heterogeneous clinical zone. So there are patients who have in-breast recurrences shortly after getting standard treatment. These are bad cancers, right? They come back after chemotherapy and while on endocrine treatment. And there the discussion is, you want to do something more, but you don’t know what more to give, because it’s a bad-apple cancer and you’re worried that you’ve already given your good chemo and they’ve already recurred on endocrine treatment. So we try to switch up the endocrine therapies there. And if there’s an obvious addition for the chemo, we think about it. Then there’s women who had chemotherapy or not 15 years ago and have a cancer in the same breast. And you tease out, is it a new primary or a late recurrence? There’s some arbitrary rules on that. But you say, “It’s been such a long disease-free interval, they don’t need anything more than just a reintroduction of endocrine therapy.” And then there’s kind of the ones in between, where you go, “CALOR suggests that there is this benefit for chemo in addition to, perhaps, endocrine therapy, but those benefits of chemo were most seen in ER-negative cancers. And many of those patients might not have had chemotherapy before, which would play into more benefit for chemo.” And that looks different than someone who had ACT 4 years ago and then recurred on an AI. So there’s a lot of discussion around the table as to what the cancer looks like. And then the other wrinkle is, apropos of today’s discussion, we often get a Recurrence Score, and now it’s 17. And they’re like, “It’s 17, but it did recur.” And then, “It recurred on an AI, so maybe it’s not so sensitive to hormones.” But then they say, “The Recurrence Score only predicts distant metastatic recurrence.” It actually does predict local recurrence, but it was optimized to predict distant recurrence. So does a local recurrence even tell you anything about how sensitive it might be? These are hard cases. And, I mean, within our group, I think it’s also fair to say that some people err more on the side of chemo and some err less on the side of chemo just based on the gestalt of to how valuable they think chemo is going to be. DR LOVE: And actually, Tari, I was just reflecting back. I know in this very room, I was interviewing Norman Wolmark from the NSABP. And he was presenting a case of a patient who had a local recurrence. And he said, “I sent an Oncotype on the patient.” And I’m like, “What!?” He goes, “Oh, yes.” Any thoughts about — I mean, it kind of, again, makes sense. DR KING: Yes, but definitely a data-free zone. I do think, though, from a surgical point of view, that it’s very important for us to get the message out to our patients and to their physicians that whether a woman saves her breast or removes her breast, that really the risk of local recurrence is the same. And we used to — the old textbooks used to say, “You can keep your breast, but you’re going to have a lifelong increased risk of in-breast recurrence if you do that.” And that’s just not true with the interplay of all the therapies that we’re giving, the systemic therapy, the radiation therapy, our decisions based on biology. Local recurrence is the same whether a woman keeps her breasts or removes her breasts. And it’s so critical as we talk about all these other treatment algorithms. Our management of the axilla is different whether someone keeps her breast or removes her breast. And so we really have lots of reasons to encourage women to keep their breasts when they can.
Relevant factors in determining whether to continue endocrine therapy beyond 5 years and the potential utility of genomic assays in this setting
DR LOVE: So we want to move on now and talk about the issue of continuing endocrine therapy beyond 5 years. And one thing we were interested in is the use of genomic assays. And globally, about half of the people in the survey — more the medical oncologists, And the other thing we asked is other clinical factors that weigh into the decision. And again, we asked the panel to rate these factors, one being original risk of recurrence at diagnosis, particularly nodal status, age, comorbidities, how the patients have tolerated adjuvant therapy and patient preference. All of these were rated very highly. Hal, can you talk a little bit about how you approach this decision and not only these clinical factors, but do you think there’s any role for genomic assays? DR BURSTEIN: Yes. So we’ve learned a lot in recent years about longer durations of antiestrogen therapy. It’s one of those things that was always staring us in the face but we weren’t paying attention to. If you look at the Oxford Overview from, like, 1990, you’ll see that there are a lot of recurrences after year 5. It’s just we didn’t pay attention to it, because we were so focused on 5 years of endocrine treatment. And in some early studies, it didn’t look like longer durations of tamoxifen was better. So now we’re in a new era, and we are really recognizing that ER-positive, HER2-negative breast cancers have that relatively long tail of risk. And the question is who to intervene with to make it better by introducing longer durations. So exactly as your survey data say, the important things here are stage at presentation and grade at presentation and then the patient experience. So a lot of older studies from MD Anderson, from Steve Chan at University of British Columbia have very nicely shown that stage and tumor grade are very powerful predictors of recurrence after 5 years of endocrine therapy. Stage III, node-positive Stage II, much higher risk than node-negative. Stage I, lower risk than Stage II and, in turn, lower risk than Stage III. So that can help tell you what the risk is of later recurrence. And then you have to talk to the patient to find out how it’s been going. And in our practice, we all have women who take these drugs and who have relatively few side effects, where it doesn’t really interfere with their day-to-day activities and they feel fine. And we have many other women who have terrible side effects and everything in between. So a lot of patients will really want to know quantitatively, as best you can, what the likely benefit of longer durations of treatment would be, and they’re weighing that against the past 5 years, where they’ve been taking these medicines and they know a lot about how they feel on these medications. DR KING: Yes, absolutely. I mean, these are very challenging decisions for the patients. Oftentimes they’ll come to the surgeon and say, “My medical oncologist told me that I could stop, but what do you think?” It’s a little bit of a security blanket. It’s just like them coming back for me to examine them each year. They think it’s going to keep their cancer from coming back. So up until this point, I’ve really based my recommendations on as how set on how they feel, what their experience has been like. Even for somebody who had high-risk disease, if they’ve just been miserable for 5 years, at some point you kind of want to give them a little bit of a break. And I think now what we’re learning more, from the more recent data, is the importance of what their original surgical therapy was. So we’re seeing now that the longer durations of therapy are actually having a big impact on rates of contralateral breast cancer. And so for women who have chosen to maintain both breasts at that initial diagnosis, reducing risk of second primary is another factor to consider when you’re deciding about whether to continue on additional therapy or not, because that’s where we’re seeing such a reduction in rates.
Extended-adjuvant therapy decision-making for a younger patient who has received 5 years of tamoxifen and tolerated it well
DR LOVE: So let’s look at some clinical scenarios specifically, starting out and thinking back. So now we’re talking about a 45-year-old woman who’s had 5 years of adjuvant endocrine therapy, in this case tamoxifen. And we put out a bunch of original risk scenarios, because we know this is a dominant issue, particularly in a patient who’s tolerating adjuvant endocrine therapy well. And you can see that in general, people continue therapy beyond 5 years for people who are node-positive. But in node-negative, you see more of a split between stopping, continuing and using a genomic assay. And again, the more common one that we’ve seen is the Breast Cancer Index. We see similar findings with the surgeon. Any comments on this spectrum, Hal, of what we see and also the issue of using genomic assays? DR BURSTEIN: So I think the data support the idea that these younger women who have had tamoxifen and had node-positive disease probably should get longer durations of antiestrogen therapy. That’s one of the sweet spots for where there really does seem to be a benefit. We know from the so-called ATLAS trial that women who’ve had 5 years of tamoxifen and pushed that out to 10 actually have a very late survival benefit. We know that node-positive disease and higher-risk cancers are more likely to get clinical benefit. And while it’s not been compared head to head, my own take on this is that the benefit of longer treatment, especially when you switch to an AI, is more pronounced if you’ve been on tamoxifen initially than on the AI initially. So the case presented here is a great setup for longer durations of treatment. In terms of the genomic decision-making, I think that most of the time I find that the patient tells you what they want to do. And you don’t need an elaborate genomic assay here. That is to say, they come in and they usually have a sense as to whether they’re in the lower-risk bucket or the higher-risk bucket. And if they’re in the lower-risk bucket and they’ve had difficulty with the medications or side effects of treatment or just don’t like being on it, they are often very comfortable. I think that the genomic data can help. And it’s now been looked at with multiple different platforms, whether it’s luminal A or B, whether it’s the BCI, whether it’s the 21-gene Recurrence Score or MammaPrint®. They all perform very similarly in terms of this prognostic information, especially for node-negative cancers. And if you feel that that information will be helpful, you can order it. In the instance of these markers for long durations of treatment, none of them have really been linked to the actual treatment intervention. It’s almost all prognostic data, which is to say after 5 years of therapy, if you had a low score, you did better than if you had a high score. The only data that have really looked at actual treatment intervention was with a derivative of the BCI looking at the MA.17 study from Paul Goss. And in that trial, actually both the BCI-low and the BCI-high groups both benefited from the longer duration of treatment. It’s just that if the score was low, there was a lower risk and, therefore, less benefit to longer durations of therapy. So this is really a prognostic marker as opposed to a marker that really tells you whether the patient will benefit from longer duration. DR LOVE: Although there are papers out there. There’s data out there that purports to show that, for example, you could identify people with a risk of recurrence between year 5 and 10 of maybe in the range of 15%. And, granted, maybe we don’t have very good treatment-related data. To me, that still seems useful compared to, say, a relapse rate of 2%, if you believe the numbers. Do you? DR KING: Yes, absolutely. I think that we, based on the original tumor characteristics, we could all define that group where we think that the risk of late recurrence is high enough to warrant continued discussions about staying on treatment. And as Hal said, sometimes patients come in with their mind made up, and other times it’s more of a discussion. But we do have to acknowledge that some of these medications are quite difficult for patients to take. And I always say I don’t want the treatment to be worse than the disease itself. So we have to take all of that into consideration. DR LOVE: Again, we went back to the panel and said, “Okay” — we presented a scenario, for example, this 40-year-old lady who now is age 45, still premenopausal, originally with a 1.5-cm node-negative tumor. And in that situation, most people would just make the decision clinically to stop therapy without a genomic assay. And yet Hal, when we said, “Okay. Assume that they’ve had a Breast Cancer Index that comes back with a high result,” and you see, again, a switch in what they would do. Any thoughts? DR BURSTEIN: So my first thought is that there’s a bit of fudge here, because my guess is they already had a genomic assay 5 years ago. DR LOVE: We said not. DR BURSTEIN: And — but if you’re giving chemo to a 1.5-cm node-negative cancer, you probably had a test that was done — or more specifically, in a premenopausal woman, if you’re not giving chemo, you probably had something to say there was a reason not to. So, I mean, in this case, again, if you tell people there’s a lot of risk and then they respond to that — and I think that makes intuitive sense — in this case, in modern series, the average risk of recurrence in years 5 through 10 is probably on the order of about 10%. That’s not a bad estimate. It’s probably a high-end estimate. Having said that, I mean, I don’t think it’s wrong decision-making. And I think that under the circumstances, she probably is at more risk than she might imagine. And so there might be a substantial benefit. If she said, “I take tamoxifen. It’s no big deal. I’m still premenopausal. I feel okay,” do I have a big problem continuing that? No. The other thing we’ve learned is from lots of studies now that there really aren’t emergent side effects that come along in years 5 through 10. It’s not that you fall off a cliff. It’s just more of the same. Those are real things, osteoporosis, bone and joint symptoms, genitourinary symptoms, sexual health symptoms, hair thinning and some of the cognitive issues people describe, but it’s the same stuff as it is in years 0 to 5 as in years 5 through 10. DR LOVE: What about the impact of age? Any thoughts in general about factoring age in as you make this decision? DR KING: Absolutely. I think that what we’ve struggled for a long time with is young age, by itself an independent prognostic factor, or is it just the biology of the breast cancers that young women tend to get? And in ER-positive disease, it seems to be an interplay between the two of them. So for the other disease types, HER2-positive disease, triple-negative disease, we’re not seeing the same impact of young age. Women, they do the same regardless of whether they’re younger or older with those disease subtypes. But with ER-positive disease, we still do see a group of younger women that do have inferior outcomes to their older counterparts. And whether that’s because they’re getting more luminal B tumors or what it is, I think there’s still a lot of work to be done. But young age is one of those things that kind of grabs us all and makes us all want to do a little bit more, and particularly the 40-year-old who’s still premenopausal at 45. I think we would all err on the side of more treatment as long as she was tolerating it.
Case: A 47-year-old woman who has completed 5 years of tamoxifen for a 1.7-cm ER/PR-positive, HER2-negative, node-negative BC and experienced minimal side effects
DR BURSTEIN: Five years ago, she had a 1.7-cm intermediate-grade tumor. At that time, we did get an Oncotype DX score. It was 17. It was low. She chose endocrine therapy with tamoxifen, has had minimal symptoms but is now having oligomenorrhea. And the question is whether she should continue to extend her endocrine treatments. She had breast-conserving surgery, so she has 2 intact breasts. She has had somewhat greater risk, therefore, for a second cancer than she might have been if she’d had a mastectomy. I think here the key factors are that she’s tolerated the medicine pretty well, again, probably because she’s still premenopausal and actually hasn’t had a lot of those side effects. And so on the one hand, I wouldn’t push her very hard to take longer durations of treatment. She has a generally good prognosis. With a score of 17, remember, her 10-year recurrence risk is still almost in the single digits for a cancer like this. But I wouldn’t object at all — and, in fact, I might tilt towards continuing for a while longer with the tamoxifen, if she was interested. I probably wouldn’t make the effort — I probably would not make the effort to shut down her ovarian function and put her on an AI, given the relatively modest benefits here. DR LOVE: Tari, it’s kind of hard for me to figure out how to interpret a 21-gene Recurrence Score at the 5-year point, in that yes, on the one hand, a low Recurrence Score shows better prognosis, but it also shows more endocrine sensitivity, usually higher ER. So maybe they would do better with more endocrine therapy. What do we know about a 21-gene Recurrence Score and 10-year risk of recurrence? DR KING: I think that’s all we do know, is that the score tells us the 10-year predicted risk of recurrence. It doesn’t with 5 years of therapy, right? It doesn’t tell us what will happen in the second 5 years with more extended therapy, with what Hal was alluding to. We have the curves that show us that some patients do have late failures, but we don’t have the data that shows that an intervention at 5 years changes that or to what extent it changes that. But coming back to the local therapy questions here or the local impact of continued antiestrogen therapy, if this patient would have had a bilateral mastectomy at the time of her original surgery, there would be even smaller benefit for her to stay on extended therapy. The fact that she’s kept both breasts suggests that she will get an incremental gain of around 1.5% risk reduction for developing a contralateral breast cancer, if she stays on it longer. When we have this discussion with women all the time at their initial diagnosis, “Should I have a lumpectomy, mastectomy, bilateral mastectomy,” and women are very fearful of their risk of contralateral breast cancer and we spend a lot of time explaining to them that in a woman particularly like this, with an ER-positive breast cancer who’s going to get at least 5 years of antiestrogen therapy, her 10-year risk of contralateral breast cancer is on the order of 3% to 5%. And so we use that. We say, “This is a very low risk that you’d get contralateral breast cancer,” and we use that as a reason to try to encourage women to preserve their breast if they can. So what the extended adjuvant therapy data are showing now is that by staying on for longer than 5 years, you’re reducing the risk of contralateral breast cancer by about 1.5%, I think, is pretty consistent across all those studies. So again, it’s a very small number, but that may be the strongest reason to keep her on it, if she has both breasts intact. DR LOVE: Kind of getting back to your case here, Hal, if she had had a Breast Cancer Index done and it showed a high result, would you have considered it? What did you actually do, incidentally, with this lady? DR BURSTEIN: So she was very comfortable continuing treatment even for what we know to be a modest benefit. It would be surprising if a different genomic test came back with a different result. There was a very nice presentation by Ivana Sestak at San Antonio, part of Mitch Dowsett’s research group. And they compared all these genomic platforms. And in the node-negatives in particular, the results were essentially overlapping. So you really shouldn’t see a discordant result. So the patients had genomic testing at baseline. They probably don’t need anything further 5 years out to refine that discussion.
Case: A 61-year-old woman who received 5 years of anastrozole for a 1.7-cm ER/PR-positive, HER2-negative, node-negative BC and now presents with a contralateral breast tumor
DR BURSTEIN: So this case is a woman who had a 1.2-cm low-grade, ER and PR-positive tumor, also node-negative, also returned at 17. She’s been on an AI, now has a contralateral breast cancer despite the AI therapy. And this one is tested and returns with an Oncotype DX of 19. So the great thing about breast cancer is a lot of people do well. The downside is that means they get second cancers and other health problems. And you have to figure those out. So this is one of those cases where the tumor arose in the setting of an aromatase inhibitor treatment. And so you have to wonder what a genomic test really tells you about the sensitivity of the tumor to an endocrine manipulation. DR LOVE: And so what’s her current situation? DR BURSTEIN: So this was a hard case, because if you look at the tumor itself, it doesn’t scream, “I need chemotherapy.” But if you look at the clinical context, which is a tumor grew during the use of an AI, it says to me, at least, that I don’t know that we can count on ongoing endocrine therapy to adequately control this. So this was a very hard discussion, but we ended up tilting towards chemotherapy here, thinking that the clinical observation trumped the genomic score. DR LOVE: Any comments? And also, the whole issue of management of second cancer. DR KING: Yes. So 1 comment, which I think we all are always cognizant of but don’t want to necessarily point the finger, but I think we all know that patients are not compliant sometimes with their medications, depending on their side effects, et cetera. So I think in this kind of situation, I’m sure that you were very careful to say, “Did you take your anastrozole? Were you taking it,” before you make that decision, because we know that people sometimes tell us they’re taking it but may not be. But certainly when a woman does develop a contralateral breast cancer, it’s very disappointing for her. She’s been on all this therapy. She thought that she was protecting herself. But essentially I tell them, “This is an independent event, and we’re just going to — we start over back at square 1, and we think about the risks from this cancer. And we make the same management decisions as we did the first time.” And again, fortunately most women are not in this situation. The rate of contralateral breast cancer is low. And we deal with it when we have to. |